Advanced materials for electromagnetic launch

Anyone can fire once.

Electromagnetic launch works, but the barrel doesn't last. Armature Systems makes the rail, pulsed-power and thermal components that determine how many times a launcher can fire.

What railgun testing has shown.

Electromagnetic launch has worked at Dahlgren, at White Sands and on the test stands of every serious naval power, but the barrel has never worked.

Rails carrying megaampere current erode with every shot. Prototype barrels needed maintenance or replacement after a few dozen shots. That's a materials problem, and the United States spent thirty years offshoring the ability to make advanced materials at the same time it offshored microelectronics.

a few dozen shots · where prototype barrels needed maintenance or replacement1,000 shots · the Phase II objective, reported met in 2018, conditions not publisheda few dozen shots · where prototype barrelsneeded maintenance or replacement1,000 shots · the Phase II objective,reported met in 2018, conditions not published
Each mark is one shot. Barrel life reached during ONR Phase I ran to tens or hundreds of shots. The Phase II objective was a thousand at ten rounds a minute.
~$500M
U.S. Navy investment in the railgun program, 2005 to 2021
Tens to hundreds
Shots of barrel life reached during ONR Phase I
120 rounds
Barrel-life target set by Japan ATLA for its 40 mm shipboard gun
~2 MA
Current at the sliding contact, where the rail wears

Sources: CRS R44175 (Apr 2022, Jan 2026); ONR Future Force vol. 2 no. 1 (2015); ATLA via Naval News, Apr 2025. Public figures, attributed to their source.

The fleet is firing million-dollar interceptors at thousand-dollar threats.

A destroyer carries 96 launch cells and can't reload underway. An electric gun adds magazine depth only if its barrel lasts long enough to empty its magazine.

96 cells. Mk 41 vertical launch cells on a Flight IIA/III destroyer. A cell that has been fired stays empty until the ship returns to port.
$0M$3M$6M$9MSM-6$10.6MPAC-3 MSE$4.3MSM-2$2.1MHVP, 5-inch guided$0.09MOne-way attack drone (threat)$0.05MEM round, unguided~$0.025Mcost per engagement, $M · published procurement figures, lot-dependentcost per engagement, $M ·published procurement figures, lot-dependent$0M$3M$6M$9MSM-6$10.6MPAC-3 MSE$4.3MSM-2$2.1MHVP, 5-inch guided$0.09MOne-way attack drone (threat)$0.05MEM round, unguided~$0.025M
Sources: CRS R44175 for railgun and HVP round costs; SASC testimony and Navy budget documents for interceptors. Unit costs vary by lot.

Three layers, starting with the rail.

In a launcher, the capacitor bank takes up the most space, and the rails wear out and have to be replaced. We make the layers that go into launchers other companies build.

rail conductor surface · layer 1pulsed-power dielectric layerthermal pathsubstrate · the integrator's partone deposition processlayers 1 to 25 µm per pass · thickness exaggerated · not to scalerail conductor surface ·layer 1pulsed-powerdielectric layerthermal pathsubstrate ·the integrator's partone deposition processlayers 1 to 25 µm per pass ·thickness exaggerated · not to scale
  • Layer 1

    The rail

    Nanostructured copper-based conductor surfaces, deposited as a graded layer rather than bonded as a plate. Figure of merit: erosion per shot, then shots per barrel.

    Now
  • Layer 2

    The pulsed-power layer

    High-permittivity multi-element dielectric layers for denser pulse-forming networks. Figure of merit: recoverable joules per liter at rated field, with breakdown strength stated.

    Roadmap, after rail data
  • Layer 3

    The thermal path

    Thermoelectric and heat-spreading layers for the barrel and the bank. Figure of merit: watts per square centimeter pumped at a stated temperature difference.

    Roadmap, after rail data

Layer descriptions are design intent. No figure of merit on this page has yet been measured on a launcher component. What has and hasn't been measured.

Where we fit in the supply chain.

We're a component supplier. Our customers are launcher integrators, and we don't sell to end users. That's why every prime is a customer and we don't compete with any of them.

Program office

Funds, tests and accepts the launcher.

  • ONR Code 33
  • NSWC Dahlgren
  • NSWC Philadelphia
  • Joint Hypersonics Transition Office
  • PEO Ships

The integrator's launcher

General Atomics EMS, BAE Systems, allied primes.

Barrel and railsArmature supplies this layer
Pulsed-power networkArmature supplies this layer
Thermal pathArmature supplies this layer
Fire control and mountThe integrator supplies this.

End user

Fields the weapon.

  • U.S. Navy
  • U.S. Army
  • Allied navies, under FMS and DDTC license

These organizations are included to describe the structure of the market. No contract, letter of intent or evaluation with any of them is implied.

Measured, not modeled.

measured ·named laboratory, method and datemodeled ·a hypothesisshots →erosion per shot ←ILLUSTRATIVE · NO DATA ON THIS CHARTmeasured ·named laboratory, method and datemodeled ·a hypothesisshots →erosion per shot ←ILLUSTRATIVE · NO DATA ON THIS CHART
This chart is illustrative. The solid line stands for a measured result and the dashed curve for a modeled one.

We publish a performance number when an independent laboratory has measured it and named the method.

There's no rail-life figure on this website. We haven't measured one on a launcher component yet, and no one else has published one. The goal of our program is to produce that number at a named laboratory, with a named method and date. Until then, every figure here is attributed to its public source, and we present our own claims as hypotheses.

See what has been measured

Railgun programs picked up again in 2025, and the barrel problem is still there.

Three programs now carry a rail-life requirement they can't yet meet with published data: the BBG(X) main gun, terrestrial batteries proposed for Guam, and Japan's destroyer gun for the 2030s.

Open the program tracker

The U.S. Navy fires the 32 MJ launcher again at White Sands, as a hypersonic test bed.

Japan fires a 5 MJ shipboard railgun at a target vessel from JS Asuka.

The BBG(X) battleship is specified with a 32 MJ railgun.

ISL (France, Germany) fires its first outdoor free-flight railgun shot.

Sources: NSWC PHD 2025 achievements; ATLA via TWZ and Naval News (2025); Breaking Defense; Naval News 6/2026.

Reading room.

We're writing plain-language articles on rail wear, cost exchange, pulsed power and how to read a performance figure.

All resources
rail cornermelt film · Al, Al₂O₃, Al-Cu
Rail physics · Coming soon

Where the rail wears: current, contact and the melt film

Every shot leaves aluminum, oxide and copper-aluminum compounds on the rail surface. The rail wears most at its corners and at the sliding contact, where the current concentrates. This article walks through the mechanism in plain terms and explains why firing a barrel once tells you almost nothing about whether it can fire a thousand times.

Read article
not publishednot publishednot publishedtarget onlynot publishednot publishednot publishednot publishednot publishedrail life disclosed
Program survey · Coming soon

Rail life disclosed: what public railgun programs publish

Muzzle energy, velocity and range make headlines. Shots per barrel almost never does. We surveyed every public electromagnetic launch program and recorded the one field that determines whether a gun is a weapon or a demonstration. For nearly all of them the answer is the same.

Read article
SM-6$10.6MPAC-3 MSE$4.3MSM-2$2.1MHVP$0.09MEM round$0.025M
Cost exchange · Coming soon

Magazine depth is a materials problem

The interceptors in a destroyer's 96 vertical launch cells cost between two and eleven million dollars each, and the ship can't reload those cells at sea. This article puts the cost-exchange numbers side by side and shows how rail life affects them.

Read article

For investors.

We make one narrow claim and show the evidence behind it. We explain the non-dilutive funding path before any conversation about equity.

Investor materials are available to qualified parties on request. Access to the data room requires an executed confidentiality agreement and, for any foreign person, prior U.S. export-control authorization. Equity in Armature Systems carries no interest in Cross Roads IP Holdings or its ventures in other sectors.